PROJECT SUMMARY Malignant gliomas collectively represent the most common primary tumors of the brain, and a particularly lethal form of cancer. The subtypes of glioma are classified according to histopathological and clinical criteria established by the World Health Organization (WHO). WHO grades II-IV gliomas are characterized by invasive growth and recalcitrance to current therapies, and thus patients with malignant gliomas currently have a very poor prognosis. Among the heterogeneous glioma cell types, one that uniquely confers resistance to radiation and chemotherapy has been identified by the expression of CD133 (Bao et al., 2006a; Liu et al., 2006). Multipotent CD133+ cells isolated from malignant glioma have the capacity to initiate and fully recapitulate disease in immunodeficient mice (Singh et al., 2004). The identification of a tumor-initiating cell type in brain tumors suggests that, like hematological malignancies, the heterogeneous cells of gliomas represent a hierarchical arrangement of differentiation states. According to this paradigm, elimination of tumor-initiating cells would provide the greatest therapeutic benefit. This prediction has prompted extensive investigation of molecular pathways that regulate tumor-initiating cells. The Hedgehog (Hh) pathway is one signaling mechanism that regulates cellular differentiation, and whose activity has been implicated in the growth of other malignancies. We have demonstrated in a large survey of patient samples that the Hh pathway is operational and activated in astrocytomas and oligodendrogliomas (WHO grades II and III) and not in primary glioblastoma multiforme (GBM; WHO grade IV) (Ehtesham et al., 2007). Within grade III gliomas, we find that expression of the Hh pathway components PTCH and GLI1 segregate to CD133+ cells. Furthermore, we have measured a significant survival advantage with pharmacological inhibition of the Hh pathway in mice bearing direct orthotopic xenografts from grade III gliomas but not from a primary GBM. We therefore hypothesize that ligand-dependant Hh signaling in CD133+ cells regulates the growth of specific types of malignant glioma. The specific aims of this proposal have been designed to rigorously test this hypothesis in a relevant preclinical model and assess the therapeutic utility of targeting the Hh pathway in malignant glioma. The in vivo studies in this proposal are innovative in the use of human glioma surgical specimens and a primary xenograft mouse model to rigorously assess the therapeutic utility of targeting the Hh pathway in established malignant gliomas. Accomplishing the aims of this study will define the types of malignant glioma and thus patients that might respond to a novel targeted molecular therapy. Information gained from these studies will also provide the foundation for optimizing therapies to target glioma cells that are otherwise resistant to current therapies.